Model Catalogue

Open and partner foundation models, served from sovereign Himalayan infrastructure.

Browse hosted inference pricing and partner-community models with transparent access terms.

CategoryProviderAvailability

Hosted inference models

Rates are loaded from the live pricing catalogue. Models marked catalogue-only are not connected to a working serving endpoint yet.

Indicative rates. Inference capacity is provisioned on request — contact us to arrange an endpoint.

LLM$0.95/hr

Llama 3.3 70B

Available on request

Large open foundation model for production text generation.

70B parameters
Text generation

Try in Playground

Deploy Endpoint

LLM$0.85/hr

Qwen 2.5 72B

Available on request

Large catalogue model for multilingual reasoning workloads.

72B parameters
Multilingual

Try in Playground

Deploy Endpoint

LLM$0.45/hr

DeepSeek-R1

Available on request

Reasoning-focused catalogue model for complex workloads.

Reasoning
Long context

Try in Playground

Deploy Endpoint

Vision$0.35/hr

FLUX.1 [dev]

Available on request

Image generation model listed for hosted inference access.

Image generation
Catalogue-only

Try in Playground

Deploy Endpoint

Partner and community models

Nepali language models — Himalaya AI Labs

These models are published by Himalaya AI Labs, a separate non-profit organisation. CloudFARM intends to host them as partner/community models; they are specialists for Nepali language, Devanagari script, and OCR rather than general-purpose peers of large 70B models.

Models by Himalaya AI Labs

LLM~5B